Subscriber access provided by UNIVERSITY OF CONNECTICUT
Communication
Diffusion-Controlled Epitaxy of Large Area Coalesced WSe Monolayers on Sapphire 2
Xiaotian Zhang, Tanushree H Choudhury, Mikhail Chubarov, Yu Xiang, Bhakti Jariwala, Fu Zhang, Nasim Alem, Gwo-Ching Wang, Joshua A. Robinson, and Joan M. Redwing Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.7b04521 • Publication Date (Web): 17 Jan 2018 Downloaded from http://pubs.acs.org on January 18, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Nano Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Nano Letters
Diffusion-Controlled Epitaxy of Large Area Coalesced WSe2 Monolayers on Sapphire Xiaotian Zhang1, Tanushree H. Choudhury2, Mikhail Chubarov2, Yu Xiang3,4, Bhakti Jariwala1, Fu Zhang1, Nasim Alem1,2, Gwo-Ching Wang3,4, Joshua A. Robinson1,2 and Joan M. Redwing1,2*
1
Department of Materials Science and Engineering, The Pennsylvania State University, University Park, PA, 16802, USA 2 2D Crystal Consortium, Materials Research Institute, The Pennsylvania State University, University Park, PA, 16802, USA 3 Department of Physics, Applied Physics & Astronomy, Rensselaer Polytechnic Institute, Troy, NY, 12180, USA 4 Center for Materials, Devices and Integrated Systems, Rensselaer Polytechnic Institute, Troy, NY, 12180, USA *E-mail:
[email protected] ABSTRACT: A multi-step diffusion-mediated process was developed to control the nucleation density, size and lateral growth rate of WSe2 domains on c-plane sapphire for the epitaxial growth of large area monolayer films by gas source chemical vapor deposition (CVD). The process consists of an initial nucleation step followed by an annealing period in H2Se to promote surface diffusion of tungsten-containing species to form oriented WSe2 islands with uniform size and controlled density. The growth conditions were then adjusted to suppress further nucleation and laterally grow the WSe2 islands to form a fully coalesced monolayer film in less than one hour. Post-growth structural characterization demonstrates that the WSe2 monolayers are single crystal and epitaxially-oriented with respect to the sapphire and contain anti-phase grain boundaries due to coalescence of 0o and 60o oriented WSe2 domains. The process also provides fundamental insights into the 2D growth mechanism. For example, the evolution of domain size and cluster density with annealing time follows a 2D ripening process, enabling an estimate of the tungsten-species surface diffusivity. The lateral growth rate of domains was found to be relatively independent of substrate temperature over the range of 700-900 oC suggesting a mass
1 ACS Paragon Plus Environment
Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
transport limited process, however, the domain shape (triangular versus truncated triangular) varied with temperature over this same range due to local variations in the Se:W adatom ratio. The results provide an important step toward atomic level control of the epitaxial growth of WSe2 monolayers in a scalable process that is suitable for large area device fabrication. KEYWORDS: tungsten diselenide, chemical vapor deposition, epitaxy, ripening, surface diffusion
2 ACS Paragon Plus Environment
Page 2 of 32
Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Nano Letters
Monolayer two-dimensional (2D) materials, particularly the family of transition metal dichalcogenides (TMDCs), have been a focus of increasing interest due to the limitations of zero-gap graphene and their own unique properties. For example, monolayer TMDCs such as MoS2 and WSe2 are direct-gap semiconductors1 with large exciton binding energies2,3 and thus are of interest for photonics and optoelectronics.4,5 In addition, WSe2 exhibits other intriguing electronic properties including native p-type conductivity,6,7 low effective mass8,9 and suitable band alignment with monolayer MoS210–12 for broken-gap tunnel field effect transistors.13 The development of device technologies based on TMDCs has, however, been hampered by difficulties in synthesizing large area monolayer and few layer films. Powder vapor transport (PVT), also referred to as powder source chemical vapor deposition (CVD), has been widely used to prepare TMDC crystal domains and films,14,15 however, it is difficult to control and modulate the source supply in this process and uniform deposition over large substrate areas is challenging. Epitaxial TMDC films have been grown on CaF2, epitaxial graphene and sapphire substrates by molecular beam epitaxy (MBE)16–21 but domain sizes are typically smaller than films grown by PVT or CVD. As a result, recent efforts have focused on the development of gas source CVD methods such as metalorganic CVD (MOCVD) for TMDCs which offers process flexibility and scalability.22,23 In gas source CVD, both the metal and chalcogen precursors are located outside of the deposition chamber which enables independent control of precursor partial pressures and ratios. MOCVD has been used to prepare large area, coalesced polycrystalline MoS2, WS2 and WSe2 monolayer and few-layer films on a variety of substrates22,24–26 and epitaxial growth of WS2 and WSe2 domains by MOCVD on epitaxial graphene have been demonstrated.23,27,28 The realization of large area single crystal TMDC monolayer films, however, requires the ability to
3 ACS Paragon Plus Environment
Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
control the density and orientation of nuclei on the substrate surface and the lateral growth rate of the domains to achieve fully-coalesced 2D monolayer films with minimal 3D growth. In this study, we demonstrate a diffusion-controlled gas source CVD process for the epitaxial growth of large area, single crystal WSe2 monolayer and few layer films on c-plane sapphire ((0001) α-Al2O3). A multi-step growth process was developed which employs modulation of the metal precursor partial pressure to independently control nucleation, domain ripening, lateral growth and film coalescence via surface diffusion processes. By separating out the nucleation and lateral growth stages, the effects of process conditions on surface diffusion and lateral growth can be clearly discerned, providing insights into the fundamental processes that control WSe2 domain growth. A schematic of the multi-step process employed in this work in shown in Figure 1 along with representative atomic force microscopy (AFM) images of the sample surface throughout the process. The WSe2 films were synthesized by gas source chemical vapor deposition (CVD) in a cold wall vertical quartz tube reactor with an inductively heated SiC-coated graphite susceptor (Figure S1). The sapphire substrate was heated to the growth temperature of 800 oC under H2 and exposed to H2Se which was held at a constant flow rate of 7 sccm throughout the entire process. Tungsten hexacarbonyl (W(CO)6) was initially introduced into the reactor at a higher flow rate (~1.2×10-3 sccm) for 30 sec to drive nucleation on the sapphire surface. Immediately after this short nucleation step, a high density of W-containing nanoscale clusters (~5 nm in size) are present on the stepped sapphire surface along with a few larger particles (Figure 1b). X-ray photoelectron spectroscopy (XPS) analysis (Figure S2) suggests that the clusters are W-rich WSex nanoparticles, consistent with results from prior cross-section TEM analysis.23 The W(CO)6 flow was then removed from the inlet gas stream and the sample was annealed in H2Se 4 ACS Paragon Plus Environment
Page 4 of 32
Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Nano Letters
for 15 min. During this annealing period, the density of nanoscale clusters decreased significantly and oriented triangular WSe2 islands with uniform size formed in a process that resembles two-dimensional ripening (Figure 1c). The W(CO)6 was then re-introduced into the inlet gas at a lower flow rate than the nucleation step (~4.2×10-4 sccm) to limit further nucleation and promote lateral growth of the domains (Figure 1d).
Figure 1. (a) Schematic diagram of the multi-step process showing variation in W(CO)6 flow rate that was used to control nucleation, ripening and lateral growth. AFM images of WSe2 grown on sapphire substrate (surface steps aligned horizontally) after (b) nucleation stage (inset shows the 5 times magnification of the surface, scale bar: 50 nm), (c) ripening stage, and (d) lateral growth stage.
The nucleation and ripening steps provide a means to control the density, size and orientation of WSe2 islands which is essential for achieving coalesced monolayer films. For example, the 5 ACS Paragon Plus Environment
Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
density of WSe2 islands on the sapphire surface after 10 mins ripening was found to increase approximately linearly from 4 to 60 µm-2 as the nucleation time was increased from 30 sec to 5 min (Figure S3) for a W(CO)6 flow rate of 1.2×10-3 sccm during the nucleation stage. The shortest nucleation time (30 sec) was chosen for subsequent studies in order to achieve the largest WSe2 domain size prior to coalescence. A detailed study of the ripening stage provides insights into surface diffusion which is the key mechanism responsible for lateral growth of 2D monolayer domains. Figure 2a shows the WSe2 surface morphology grown using a constant nucleation time of 30 sec and increasing ripening times ranging from 7.5 min to 30 min. As ripening proceeds, triangular domains start to appear with increasing density and size while the small clusters between the triangles decrease in
Figure 2. (a) AFM images of WSe2 grown on sapphire substrate under 30 sec nucleation time and varied ripening time for 7.5, 15, and 30 min. (b) WSe2 domain size as a function of ripening time and cubic root of ripening time (inset). (c) Cluster density as a function of ripening time and reciprocal of ripening time (inset). (d) Substrate surface coverage as a function of ripening time.
6 ACS Paragon Plus Environment
Page 6 of 32
Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Nano Letters
number. The size of the triangular domains (measured by the edge length of the triangle) increases rapidly up to ~200 nm within the first 15 min of ripening (Figure 2b) with a corresponding decrease in cluster density (Figure 2c). During this same time period, the total surface coverage decreases from 15% to 8.5% (Figure 2d) mainly due to differences in the projected areas of clusters versus triangular WSe2 domains. Beyond 15 min, the domain size, cluster density and coverage remain relatively constant or decrease only slightly with further annealing time. The time dependencies of the domain size and cluster density within the first 15 min are consistent with classical models of 2D ripening29–32 whereby large particles increase in size as t1/3 (Figure 2b inset) and small particles decrease in density as 1/t (Figure 2c inset). The exact mechanism by which ripening occurs is not known although it is believed to be associated with surface diffusion of W adatoms or migration of W-rich WSex clusters. A simplified comparison of the vapor pressures of W (
